The HERMES Study
Research type
Research Study
Full title
Tele-ophthalmology- enabled and Artificial Intelligence-ready referral pathway for community optometry referrals or retinal disease: a cluster Randomised Superiority trial with a linked Observational Diagnostic Accuracy Study
IRAS ID
285992
Contact name
Konstantinos Balaskas
Contact email
Sponsor organisation
Moorfields Eye Hospital NHS Foundation Trust
ISRCTN Number
ISCTN18106677
Duration of Study in the UK
2 years, 10 months, 31 days
Research summary
Early diagnosis of retinal disease depends on the imaging technology called Optical Coherence Tomography (OCT). OCT is increasingly available in high street optometrist practices but is not accompanied by hospital-level expertise in interpreting these scans. This is leading to a large number of patients being referred inappropriately, increasing the pressure on hospital eye services and delaying access to care for patients that do need treatment.
This study aims to use two forms of new technologies to improve this referral process, by helping hospitals and high street optometrists to work together to refer the ‘right patient at the right time’.
The first technology we will evaluate is ‘tele-ophthalmology’ in which OCT scans taken by high street optometrists are reviewed by hospital specialists remotely. High street optometrists with OCT will be divided into two groups: half of the practices, selected by chance, will continue to refer patients using the existing paper-based system, with the other half installing a leading tele-ophthalmology platform (‘Big Picture’) to allow instant transfer of scans to the eye hospitals for review and advice within 24 hours. We will involve 588 patients overall in this part of the study.
We can then assess whether the new referral system can safely lead to fewer unnecessary visits to eye hospitals and improves the time it takes for patients referred to be seen or treated.The second technology we will evaluate is interpretation of retinal scans by ‘Artificial Intelligence’ (AI) using ‘the Moorfields-DeepMind algorithm’ on all OCT scans collected from our participating high street optometrists and asses the accuracy in providing the correct advice for referral. This exciting development could enable expert-level care to be digitally embedded into this referral pathway but first we need further evidence of how this AI technology would perform in the real world.
We will also assess the cost of the new system and collect feedback from patients and health practitioners about it -confidence in its safety and data privacy, effect on patient experience and the potential role of Artificial intelligence for eye referrals.
Summary of results
The HERMES study assessed the effectiveness of a teleophthalmology referral pathway between community optometrists and hospital eye settings. Teleophthalmology is the review of medical information which has been electronically exchanged. Using this technology, referrals with eye scans from community optometrists were remotely reviewed by a hospitalbased eye specialist.
294 participants were recruited by 26 optometry sites, of which 158 participants were referred via the teleophthalmology referral platform, and 136 participants were referred via the standard referral pathway. The teleophthalmology pathway was effective and reduced the proportion of unnecessary urgent referrals by almost 60%, it decreased the proportion of incorrect referral urgency by 25%, significantly reduced the proportion of incorrect diagnoses and if implemented, is likely to have lower costs and greater effectiveness.
The role of Artificial Intelligence to improve hospital referrals was also assessed. Artificial Intelligence is a computer programme that is trained to do tasks which require human intelligence. Here we used an Artificial Intelligence model that can look at eye scans and recommend if a hospital referral is needed or not. The model was used on suitable scans from study participants. We then compared the model to optometrists and hospital experts for deciding if a hospital referral was required. The model sometimes made different referral decisions than hospital experts, but it made similar decisions to optometrists.REC name
London - Bromley Research Ethics Committee
REC reference
20/LO/1299
Date of REC Opinion
26 Jan 2021
REC opinion
Further Information Favourable Opinion